Distribution-free inference in record series

被引：0

作者：

Anis S. Hoayek

Gilles R. Ducharme

Zaher Khraibani

机构：

[1] Université de Montpellier,Institut Alexander Grothendieck

[2] Université Libanaise,undefined

来源：

Extremes | 2017年 / 20卷

关键词：

Estimation; Goodness-of-fit test; Linear drift model; Records; Confidence interval; Yang-Nevzorov model; Primary–62N02; 62G32; Secondary–62F12; 62F10;

D O I：

暂无

中图分类号：

学科分类号：

摘要：

Let {Xt, t = 1} be a time series. A (upper) record is a value Xj such that Xj> max{X1,…,Xj-1}. Some popular models in record theory are the Yang-Nevzorov and the Linear Drift models. The stochastic behavior of records under these models has been much studied and many interesting distribution-free properties have been unearthed. However the estimation of the parameters of these models has been less explored. This work introduces some estimators of these parameters. Their behavior is investigated theoretically and by numerical simulations. It is shown that some of these estimators are easy to compute and their asymptotic properties are accurate and distribution-free. Some goodness-of-fit tests for these models are also presented.

引用

页码：585 / 603

页数：18

共 25 条

[1]

Ballerini R(1985)Records from improving populations J. Appl. Probab. 22 487-502

[2]

Resnick S(1987)Embedding sequences of successive maxima in extremal processes, with applications J. Appl. Probab. 24 827-837

[3]

Ballerini R(1999)On records and related processes for sequences with trends J. Appl. Probab. 36 668-681

[4]

Resnick SI(1993)Parametric likelihood inference for record breaking problems Biometrika 80 507-515

[5]

Borovkov K(1996)On distribution-free inference for record-value data with trend Ann. Stat. 24 2655-2678

[6]

Carlin BP(1996)Properly rescaled components of smooth tests of fit are diagnostic Aust. J. Stat. 38 61-74

[7]

Gelfand AE(2015)A non parametric exact test based on the number of records for an early detection of emerging events: Illustration in epidemiology Commun. Stat. Theory Methods 44 726-749

[8]

Feuerverger A(2016)On the parametric maximum likelihood estimator for independent but non-identically distributed observations with application to truncated data J. Stat. Theory Appl. 15 96-107

[9]

Hall P(1988)Forecasting records by maximum likelihood J. Am. Stat. Assoc. 83 331-338

[10]

Henze N(2014)Modeling record-breaking stock prices Phys. A: Stat. Mech. Appl. 396 114-133

← 1 2 3 →